Ultrastructural localization of the neurotrophin receptor (TrkA) in cultured rat pheochromocytoma PC12 Cells: three-dimensional image analysis by high voltage electron microscopy.

Nerve growth factor (NGF) is a well-known neurotrophic factor and the NGF signaling through the receptor, TrkA, plays important roles in regulating neuronal differentiation and survival. A recent study has demonstrated that the TrkAs expressed in undifferentiated PC12 cells were associated with caveolae, which were invaginated small pits on the plasma membrane. Caveolae are frequently seen in many cell types such as endothelial cells, fibroblasts and hepatocytes, but few in neurons. In the present study, we performed immunocytochemistry of TrkA in differentiated PC12 cells and analyzed the ultrastructural localization of TrkA by conventional electron microscopy and high-voltage electron microscopic (HVEM) tomography. The TrkA immunoreactivities were mainly associated with the cytoplasmic vesicles (10-30 nm in diameter) and a part of the plasma membrane. The HVEM tomography showed that the TrkA immunoreactivities were often assembled into ring-like structures (400-800 nm in diameter) near the plasma membrane, unlike typical flask-shaped invaginations of caveolae (50-100 nm in diameter). These results suggest that TrkA are not localized in the caveolae, at least in differentiated PC12 cells, but other invaginations are involved in a novel process of internalization of ligand-bound TrkA.

TrkA immunoreactive astrocytes in dendritic fields of the hippocampal formation across estrous.

Neurotrophins are important modulators of structural synaptic plasticity. (Through trophic action (Jordan. J Neurobiol 40:434-445, 1999), astrocytes serve as permissive substrates to support axonal regrowth (Ridet et al. Trends Neurosci 20:570-571, 1997), and are involved in estrogen-induced synaptic structural plasticity (Garcia-Segura et al. Cell Mol Neurobiol 16:225-237, 1996). Previously, we reported that tyrosine kinase A receptor (TrkA) immunoreactivity was present both in presynaptic neuronal processes (axons and terminals) and in select astrocytes of the male rat hippocampal formation (Barker-Gibb et al. J Comp Neurol 430:182-199, 2001). We show that the number of TrkA-immunoreactive astrocytes in female rats fluctuates 16-fold across the estrous cycle in dendritic fields of the hippocampal formation, with the greatest number at estrus after the peak plasma estradiol concentration of proestrus. Few TrkA-labeled astrocytes were found in ovariectomized animals; after estrogen replacement, this number increased by 12-fold in the hippocampal formation, indicating estrogen-mediated induction. Dual-labeling studies showed that TrkA-labeled astrocytes were also immunoreactive for vimentin, a protein expressed by reactive astrocytes. Ultrastructural analysis of the dentate gyrus molecular layer demonstrated that TrkA immunoreactive astrocytes are positioned primarily next to dendrites and unmyelinated axons. Because nerve growth factor (NGF) has been reported to stimulate astrocytes to function as substrates for axon growth (Kawaja and Gage. Neuron 7:1019-1030, 1991), these findings are consistent with the theory that TrkA immunoreactive astrocytes serve a role in structural plasticity, axon guidance, and synaptic regeneration across the estrous cycle in the hippocampal formation.

Morphological variation in the tyrosine receptor kinase A-immunoreactive periodontal ligament epithelium of developing and mature rats.

Tyrosine receptor kinase A (trkA) is the high-affinity receptor for nerve growth factor. It has been found in several non-neuronal cell types, indicating biological roles independent of neural function, as well as in the nervous system. An initial study demonstrated that an antibody to the full extracellular domain did not label periodontal ligament epithelium (PLE; also known as epithelial rests of Malassez), but that another antibody which recognises a truncated 41-kDa form of trkA did label PLE. Thus, truncated trkA-immunoreactive (-IR) PLE was further investigated here in developing molars of young rats, and in its mature form in adult rat molars, for its reaction to moderate or deep molar injuries, and for its appearance along the continuously erupting incisors of mature rats. In some of the adult rat molars we also analysed the association of nerve fibres with PLE using antibodies for p75 neurotrophin receptor or peripherin. Rat jaws were fixed with 4% formaldehyde and demineralised, and bound antibody was detected with avidin-biotin-peroxidase and diaminobenzidine or fluorescence procedures. Light microscopy showed great variation in the appearance of trkA-IR PLE and considerable morphological changes during the eruption of molars and incisors. By electron microscopy it was shown that trkA-IR was not uniformly distributed in PLE cells but rather was concentrated in the peripheral zones of each cell cluster. Tooth injury did not influence the form or occurrence of PLE unless there was specific destruction of a ligament region. Qualitative analyses of nerve fibres showed that they only rarely innervated PLE in adult rats, indicating that the truncated receptor has non-neuronal functions in this epithelium. These results suggest that neurotrophin growth factors, acting via truncated trkA receptors, affect the interactions between PLE cells and the periodontal ligament, with fewer PLE interactions with nerves. Furthermore, the expression of these receptors on PLE supports the possibility that these cells are active during tooth development and eruption rather than being merely passive remnants of the degenerating tooth sheath. The similar trkA-IR of PLE and junctional epithelium, as well as their structural association, suggests interactions between these two epithelia.

Developmental dependency of Meissner corpuscles on trkB but not trkA or trkC.

The distribution of S100-immunoreactive (ir) corpuscular endings was examined in the palate of wildtype and knockout mice for trkA, trkB or trkC. In wildtype mice, S100-ir corpuscular endings were abundant at the top of palatal rugae. The endings contained 2-4 parallel arrays of S100-ir neurites. The distribution of S100-ir nerve endings in trkA and trkC knockout mice was similar to that in wildtype mice; S100-ir corpuscular endings were abundant in palates of the mutant mice. In trkB knockout mice, the palate was devoid of corpuscular endings, An immunoelectron microscopic method indicated that S100-ir corpuscular endings were identical to Meissner corpuscles. The normal development of Meissner corpuscles is probably dependent on trkB but not trkA or trkC.